Learn Before
Critiquing a Prompt for a Custom Extraction Task
A developer is trying to use a large language model to extract specific information from product descriptions. They are getting inconsistent and poorly formatted results. Based on the case study below, critique the developer's initial approach and recommend a specific change to the prompt that would likely improve the model's performance. Justify your recommendation by explaining the principle behind it.
0
1
Tags
Ch.1 Pre-training - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Ch.2 Generative Models - Foundations of Large Language Models
Ch.3 Prompting - Foundations of Large Language Models
Evaluation in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Examples of Few-Shot Learning Applications in NLP
Enabling Few-Shot Learning with Multiple Demonstrations
Input-Output Patterns in Few-Shot Learning
Sufficiency of Demonstrations in Few-Shot Learning
Applying Few-Shot Learning to Complex Reasoning Tasks
A user provides the following text to a large language model to get it to classify movie reviews:
Review: The plot was predictable and the acting was wooden. I was bored the entire time. Sentiment: Negative
Review: An absolute masterpiece! The cinematography was stunning and the story was deeply moving. Sentiment: Positive
Review: It was a decent film. Not the best I've seen this year, but it had some good moments. Sentiment: Neutral
Review: I couldn't stop laughing from beginning to end. A brilliant comedy. Sentiment:
The model correctly responds with "Positive". Which statement best analyzes the primary reason for the model's successful performance on this task?
Constructing a Few-Shot Prompt for a Novel Task
Critiquing a Prompt for a Custom Extraction Task
Example of a Few-Shot Prompt for Polarity Classification